Below are the tasks we worked on during our last sprint.

Regarding Software Factory:

Regarding our contributions to Zuul and Nodepool:

  • We've updated the fix to nodepool that allows the node launch process to continue if a newly spawned node ends up in failing state but cannot be removed. The fix is now ready for review:
  • We continued working on react-zuul web interface using a better redux store design. We added follow-up page for jobs, build and config-errors notification drawer.
  • We investigated using prometheus to monitor zuul metrics and trigger alert when queue gets too long:
  • We contributed a patch to zuul-jobs for fixing sphinx job to run w/o root access
  • bridge.o.o is online and running only ansible, part of our cfg-mgmt-update topic. As well, base playbooks are all ansible, no more puppet.
  • Coming week at PTG is to focus on removing more puppet in favor of ansible.